© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 7246
    Prev
    Next
    INDEX
    Explanations

    actions and roles related to job responsibilities and tasks performed in various professions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    194
    +0.12
    0.6%
    588
    +0.11
    0.5%
    25
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.12
    0.12
    588
    +0.11
    0.07
    437
    +0.10
    0.04
    Negative Logits
    Myth
    -0.74
     clich
    -0.70
     hypocrisy
    -0.67
    Reason
    -0.65
    natureconservancy
    -0.64
    isSpecialOrderable
    -0.63
     unworthy
    -0.63
     absurdity
    -0.62
    monkey
    -0.62
     ignorance
    -0.62
    POSITIVE LOGITS
     repairing
    1.20
     repair
    1.13
     repairs
    1.10
     transporting
    1.08
     overseeing
    1.06
     assisting
    1.04
     servicing
    1.04
     administering
    0.98
    repair
    0.96
     cleaning
    0.94
    Activations Density 1.823%

    No Known Activations